It's simple, really: just assume whatever you want to be true, and you've got your conclusions.

Divorce and out-of-wedlock childbearing cost U.S. taxpayers more than $112 billion a year, according to a study commissioned by four groups advocating more government action to bolster marriages.

...

Reducing these costs, [study author Ben] Scafidi said, "is a legitimate concern of government, policymakers and legislators."

And:

"The study documents for the first time that divorce and unwed childbearing — besides being bad for children — are costing taxpayers a ton of money," said David Blankenhorn, president of the Institute for American Values.

"We keep hearing this from state legislators, 'Explain to me why this is any of my business? Aren't these private matters?'" Blankenhorn said. "Take a look at these numbers and tell us if you still have any doubt."

Tell you if we still have any doubt? Okay.

Dear Mssrs. Scafidi and Blankenhorn,

Please consider this your official notification of our doubt.

Most sincerely,
Rational People.

Seriously — doubt? Just for starters, we'd like to know how you came up with your unbelievable statistics.

Scafidi's calculations were based on the assumption that households headed by a single female have relatively high poverty rates, leading to higher spending on welfare, health care, criminal justice and education for those raised in the disadvantaged homes.

Ah, now we understand. So it's not really a study so much as it's a baseless assumption with numbers wrapped around it. Gotcha. Now it makes sense, because we know it's utter crap.

This fucking study doesn't even bother to establish causality. It seems content to assume that it's the divorces and the unwed parents that lead to more poverty, but it ignores the fact that it's just as valid (or more so) to assume that the poverty leads to increased rates of divorce and single-parenthood. And if poverty is the problem that causes this massive $112 billion issue, then why not combat that? Spending more money on increased employment, job training, and economic recovery are all general examples that might work a damn bit better than trying to combat poverty by offering marriage skills "training" on the taxpayer's dime. (That's all of us, folks.)

Perhaps — and we're just spitballing here — perhaps poverty is the root cause of all of the problems here. The study agrees that poverty leads to increased spending on criminal justice, welfare, health care, and education...but then doesn't see that the same poverty, along with those particular corollaries, is the most likely cause for unwanted pregnancies leading to single parenting and unhappy marriages ending in divorce? The term "intentionally fucking obtuse" leaps to mind here. Here's one of their "striking examples":

[Elizabeth Ananat and Guy Michaels] are able to study married couples who do and do not divorce and conclude that "divorce significantly increases the odds that a woman with children is poor."

And this suggests causality how? The fact that a divorced woman is more likely to be poor doesn't establish the cause of that poverty in any way, shape, or form. In fact, we'll stick to our hypothesis that it's more likely that the poverty causes the divorce, rather than the other way around. We don't think that's any more statistically valid, off the cuff, but it makes a whole lot more sense in context. And we didn't even have to make up any math to offer our idea!

It's telling that the Scafidi paper, while putting great stock in the Ananat-Michaels results, entirely ignores the actual point of that data.

One of the primary points of the Ananat-Michaels results was that the mean income does not decrease due to divorce. Using specific indicators that endeavor to isolate the relationship between divorce and poverty from any other factors (though we're not sure that we agree they were totally successful), Ananat and Michaels demonstrated that, statistically, many divorced women will earn less, but that those who end up earning more will earn substantially more, which (apparently) more than offsets the poverty-stricken women when it comes to determining the mean income.

In other words, since it is the overall income of these women, collectively, that will affect other taxpayers (like you and us), Scafidi's conclusion is utter bullshit. Yes, the individual women themselves will often end up in poverty, and we can (and should) certainly be concerned about that, but this nonsense about our taxes having to support scads of unmarried mothers is just Scafidi's imagination, inspired by a healthy dose of, er, sponsorship from the Institute for American Values.

As Justin Wolfers explains:

The Ananat-Michaels result is that divorce seems to help the finances of about as many women as it hurts, and those who gain, may gain more than those who lose. But this report counts up the costs to the taxpayer from the women who lose income, but refuses to count even a single dollar of the rise in taxes linked to those who gain income.

What incredibly selective statistics you've discovered, Benjamin Scafidi! And all it took was a bunch of people with an agenda asking you to write up some numbers for them to make their case look good. Nice job, you fucking weasel.

Wolfers continues:

Amazingly, the advocates [Scafidi and his fundamentalist overlords] put together "fiscal" costs of divorce without even understanding the tax code. The U.S. tax system is structured so that when poor single mothers marry men with higher incomes, in most cases, the total tax paid by husband and wife would fall. Yet this isn't counted.

Those poor single women aren't robbing us of tax revenue, they are actually paying more than if they were married! (Yes, the tax code does include a marriage penalty for some couples who are both high earners, but for most couples, the U.S. gives you a tax break for getting married.)

Please do have a look at the rest of Wolfers' article — it's pretty incredible how intellectually dishonest Scafidi's piece of shit paper is.

And there's more to consider: if you truly do believe that divorce and unwed or single parenting lead to more poverty, and that therefore the government has a vested interest in lessening the impact of that poverty, then it seems to us that a rather logical conclusion is that we should focus on teaching the facts about contraception in public schools. All this bullshit "abstinence only" nonsense is only making the likelihood of unwanted (and, as such, usually unwed) pregnancies even higher. In addition, when two teens who weren't planning on getting married any time soon (if at all) get married due to an unplanned pregnancy, that certainly isn't the best bet for a long, happy marriage. We're willing to bet the statistics on that assertion hold up better than Scafidi's lopsided math.

But wait — there's something else that might be even more effective in confronting this problem. The issue here seems to be the effects of poverty on children who are raised by single parents or who experience a divorce. It strikes us that, if only we had some way to terminate unwanted pregnancies...some way to stop a pregnancy before a zygote developed into an actual baby, perhaps some kind of, we don't know, abortive procedure of some sort...then we'd have a whole lot less single parents and divorced parents (since we'd have less parents in general, and fewer resultant forced or reluctant marriages, which are inherently less likely to last). So perhaps, in addition to adequate sex education, the government should also promote abortions as a viable alternative to having a baby, when prospective parents really just aren't ready to have one from a social and economic perspective. Having a baby when you aren't ready to support one leads to — you guessed it — a higher likelihood of poverty. We aren't saying that abortions are a wonderful alternative, or that they are for everyone (though the same can be said of divorce and single parenthood, and of marriage itself), but we are saying that educating people about the facts of contraception and abortion would very likely decrease the number of unexpected pregnancies, unwanted marriages, and broken families.

In case you'd like to take a closer look at Scafidi's very scholarly "study," you can head over to the Institute for American Values where it's housed in their "Center for Marriage and Families" and listed under "real, honest-to-gosh scientific-like research." Try your best not to be sidetracked by their ham-handed arguments against same-sex marriage, and their diatribes about the "proper" roles of men and women in a "proper" family.

What a bunch of fucking disingenuous assholes.

You know what? Here's our idea for a "study." We think Fundamentalist Christianity is fucking up our world. Further, we'll assume that Fundy families are more likely to be stupid, and as such, Christianity leads to more taxes being spent to educate their indoctrinated kids who keep getting left back over and over again because they think that early man lived side-by-side with fucking dinosaurs and rode them like bucking fucking brontos. Also, of course, more Fundamentalist-educated nitwits are ill-informed (or outright misinformed) about birth control than any other group (thanks to abstinence-only sex "education"), which leads to more unwanted teen pregnancies — which, due to the indoctrinated religious idiocy surrounding abortion, are invariably carried to term and often result in higher incidences of low-income teens getting married too young (since good "religious values" demand that they do) and earning too little (since they were so badly and briefly educated that they can't make their way in the professional world)...which, in turn, results in more poverty. More poverty, as we all know (we just read a study about this, if we recall), drives increased spending on welfare, health care, criminal justice, and education for those raised in such a stupidly disadvantaged home.

Of course, none of these assumptions are verified — we just reckon it's all true. But the obvious conclusion is that the government has a vested interest in destroying Fundamentalist Christianity. How could you come to any other conclusion? So let's get on that, and institute a government program using our tax dollars and yours to wipe out Fundamentalist Christianity in our country before it ruins our economy and quality of life. Heck, just to be safe, let's wipe out all Christianity. Or better yet, all religion. That oughta do it! After all, given the fact that the overwhelming majority of Americans are Christians (and all but a small handful are associated with some religion), we've just proven, beyond any doubt, that religion is costing America way more than $112 billion per year in tax money.

Reducing these costs by eradicating religion, the Two Percent Company said, "is a legitimate concern of government, policymakers and legislators."

At this point, we could either say "Oh, wouldn't it be nice?" or "You know how fucking facetiously we're writing this, right?" But it wouldn't matter; quote mining is as quote mining does, or so they say.

Fucking disingenuous assholes.

(It bore repeating.)

And how does one go about eradicating religion?

"And how does one go about eradicating religion?"

The same way one goes about eradicating "divorce and out-of-wedlock childbearing" I imagine.

Nice new commenting system by the way. I like it.

Please see comment #13 to the Justin Wolfers post at Freakonomics.

Also, you may want to actually read that report--including the footnotes--if you truly care about how we came up with our estimates.

Can you be a little more specific on which post? Justin Wolfers has written fifteen different posts.

Fucking disingenuous arseholes is far too polite. How about a study to determine how much religious tax exemptions are costing the American tax payer. Something tells me those tax breaks aren't coming cheap; it's not like they're keeping to their side of the bargain and keeping their noses out of politics after all.

As requested, Ben Scafidi, we've taken the time to read Comment Number Thirteen over on Justin Wolfers' thread.

Now do you mind if we ask you: why?

It's an eloquent, lucid, and rational response to Justin's piece — whether we think you've "proven your point" or not (we don't). However, your response is written to Wolfers' piece, and really doesn't address our points at all. Zip, kid — nada. So we're entirely unsure as to how we're supposed to suddenly "see the light" of the Scafidi Economic Perspective when our arguments against it haven't really been addressed. You are, of course, under no obligation to respond to our arguments, but since you chose to come to our site and leave a comment, we'd ask that, in the future, you make an effort to ensure your comments here address our arguments instead of directing us to a comment on another site that is almost entirely meaningless in the context of what we wrote.

In brief, we only addressed one mathematical concern with your work, and that lay in the fact that the Ananat-Michaels findings suggest (in line with pretty obvious common sense) that divorce has an effect on the financial status of individual women and families, but that the overall economy — the one that would affect the taxes the rest of us have to pay — evens out quite nicely. Yes, Ananat-Michaels were arguing against the hypothesis that divorce doesn't tend to negatively effect individual financial situations, but in doing do, they put forth very clearly that the mean income — the overall financial situation of the general economy — is not affected by the factors in their research. You're quite welcome to continue to claim that this "supports" your theory that divorce in general is a burden on taxpayers in general, but your reasoning simply doesn't hold up.

However, there are at least three other important points to consider, Ben.

One: as we implied, we don't remotely agree that the Ananat-Michaels methodology does anything to help them reach sound conclusions in the first place. They are attempting to isolate the causal factors surrounding divorce, so that they can see if any kind of causation flows from divorce to individual financial status. The instrumental variable they've chosen, however — the gender of the first-born child — is nothing but an arbitrary mathematical construct. It suggests nothing, and certainly doesn't take the reverse causation (from individual financial status to divorce) out of the picture in any effective way.

You see, the idea that an observed statistical difference in the trend of first-born genders (especially a difference so small) can hold any particularly effective predictive powers is purely absurd. And presuming that this factor a priori excludes myriad other factors is a pretty big presumption. We don't agree that it does exclude other factors, and we don't think that any kind of isolation has been achieved with this instrumental variable.

Of course, that's simply because of the second important point, Ben: not to throw your entire life's work in the garbage, but economics is not a mathematical science. And this is a problem, because economists, and economics scholars, like to think that it is. We entirely disagree with the folks cheering Justin Wolfers on (over on his Freakonomics thread) for "proving" that marriage and divorce have nothing to do with economics. Of course they do. We would never suggest differently.

But the reason why marriage and divorce are such a big deal, and can throw a giant wrench into economic systems, is because economics is really an anthropological science, but one in which attempts are made to bind phenomena to observed demographics and statistics so that one can pretend to be doing effective, useful, and predictive math.

Which is bullshit.

It hasn't escaped our notice that, in the field of economics, paper after paper, paragraph after paragraph, theorem after theorem...they're almost invariably prefaced with something along the lines of: "In a closed system, with two consumers, six suppliers, all equidistant from each other, with exactly the same product, of exactly the same quality..." and so on. The fact that so many conditions must be introduced in order to get "meaningful" mathematical results in economics means, frankly, that they just aren't meaningful — and they certainly aren't predictive, since the only thing they'll predict are the results of an identical system to the one used in order to derive the original math, and such rigidly defined systems are going to be awfully hard (read: virutally impossible) to come by in the real world. So what's the fucking point?

Hey, economists can be useful, sure. It's entertaining to watch them argue minute and irrelevant points. And, more seriously, it helps to understand economic models, even if that understanding won't actually let you predict, with any certainty, what's going to come next in any real-world economic system. Our point here, Ben, is that trying to pretend that economics is a science grounded in math instead of one grounded in anthropology (and/or sociology, and/or psychology) renders an analysis based on that flawed premise pretty darn useless.

Our third point, though, is what we come down to. We don't care about your numbers, Ben — we don't believe they prove your point, we don't believe your conclusions are accurate or even accurately derived in a real-world system, and we don't believe you did your work honestly and objectively in the first place...but that's not really our point.

Our point is, quite simply, that you — and David Blankenhorn — are quite simply and utterly wrong. There are many things taxpayers should be concerned about — the holistic effect of a thousand other people's divorces isn't one of them. The dishonesty comes in claiming that this is a taxpayers' concern, and it simply isn't. It's a "family values" concern, which is, of course, a code phrase for Fundamentalist Christian concern, these days. They care, and they'll use any dishonest means necessary to make it seem more important — and more causal — than it actually is.

As we said, the precise same reasoning holds up as well in our proposed eradication of religion as in the Fundamentalists' proposed eradication of divorce (even better, in our opinion) — but ours was tongue-in-cheek. Even better, John Wilson brings up the fact that tax exemptions for religious organizations are of far greater concern to taxpayers than the aggregate effect of other people's divorces — and those exemptions are granted on conditions that are clearly not being met today (and, we'll guess with a high level of certainty, never really were being met since the ratification of the Bill of Rights).

Who's really fucking whom, here? People like us, who don't want to interfere with other people's private decisions about marriage and divorce, because we do not believe that these individuals are impacting our own financial status in any meaningful way? Or people in support of religious organizations, who not only want to interfere with other people's private and personal decisions, but are getting a tax-free ride off the government while breaking their agreement to not get involved in government in the first place?

Freedom of religion is a great thing. But the tax laws surrounding it are fucking silly — by granting religious organizations tax exemption, the government is granting religion special status, which seems to fly in the face of the First Amendment. By trying to stick to the First Amendment, we've actually managed to piss all over it. This may have been fine way back when the atheists were quieter and unnoticed, and people could pretend that freedom of religion did not mean that we are also free to practice no religion. But with over 13% of Americans checking off "atheist" on the census reports nowadays, it's time to stop pretending. By granting religious practitioners the special status of tax exemption, the government discriminates against religious non-practitioners. It's an excellent illustration of the old "non-stamp-collector" axiom, incidentally — if one could really lump all of us non-believers into one group, and the churchies so often claim that "atheism is just another religion," then why can't we form a tax-exempt organization? We can't, of course, because when it comes to that kind of thing, the believers are all too content to agree that we aren't really any kind of group, religious or otherwise. So we don't get tax-exempt status...and they do. And we have to keep paying taxes that support infrastructure that keeps them warm and cozy, but they don't do jack shit for us. That's not exactly a just, egalitarian, liberated democratic republic. Or would you disagree?

So the question becomes: why would you concentrate on this exceedingly disingenuous "divorce is fucking the taxpayers" trail, Ben, when you could be concentrating on the far more demonstrable, far more accurate, far more signficant, and far more important "religious tax exemption is discriminatory and fucking the taxpayers" trail?

And the simple, obvious answer: because you weren't paid to. Instead, you were paid by a special interest group to construct a case to support a specific pre-conceived conclusion. Which, we believe, was indicated by the fucking title of our post to begin with.

Keep on keepin' on, weasel.

I take a little umbrage at this particular block:

Of course, that's simply because of the second important point, Ben: not to throw your entire life's work in the garbage, but economics is not a mathematical science. And this is a problem, because economists, and economics scholars, like to think that it is. We entirely disagree with the folks cheering Justin Wolfers on (over on his Freakonomics thread) for "proving" that marriage and divorce have nothing to do with economics. Of course they do. We would never suggest differently.

But the reason why marriage and divorce are such a big deal, and can throw a giant wrench into economic systems, is because economics is really an anthropological science, but one in which attempts are made to bind phenomena to observed demographics and statistics so that one can pretend to be doing effective, useful, and predictive math.

Which is bullshit.

It hasn't escaped our notice that, in the field of economics, paper after paper, paragraph after paragraph, theorem after theorem...they're almost invariably prefaced with something along the lines of: "In a closed system, with two consumers, six suppliers, all equidistant from each other, with exactly the same product, of exactly the same quality..." and so on. The fact that so many conditions must be introduced in order to get "meaningful" mathematical results in economics means, frankly, that they just aren't meaningful — and they certainly aren't predictive, since the only thing they'll predict are the results of an identical system to the one used in order to derive the original math, and such rigidly defined systems are going to be awfully hard (read: virutally impossible) to come by in the real world. So what's the fucking point?

Hey, economists can be useful, sure. It's entertaining to watch them argue minute and irrelevant points. And, more seriously, it helps to understand economic models, even if that understanding won't actually let you predict, with any certainty, what's going to come next in any real-world economic system. Our point here, Ben, is that trying to pretend that economics is a science grounded in math instead of one grounded in anthropology (and/or sociology, and/or psychology) renders an analysis based on that flawed premise pretty darn useless.
You aren't beating merely on economics, you're trashing modeling of all sorts. Yes, there are many conditions in models which are not exactly present in reality, but that's the point. You use less detail in order to keep things simple and derive useful insights. What is meant by "in theory" is "with certain details omitted"; those details may or may not be important but therein lies the difference between good modeling and bad. Yes, it is difficult to figure out what's going to happen next, but predicting the future in any complex system is never a simple or easy task.

I like Al Roth's response to Amos Tversky on a similar point.

PS. Nice comment system.

I don't think your assessment is entirely fair, "Pareto." Models, and modeling, can be useful, provided you don't lose sight of the fact that they are models, and thus imperfect representations of reality at best.

Economics, like the 2% guys point out (rightly, in my opinion) is a social science with pretensions of being a "hard" science, and I think this is a reason that economists in particular have a tendency to overlook all the various conditional factors ("assume a closed society with perfect competition"-type things) to which their models are subject and subsequently mistake them for reality.

A semi-serious hypothesis I toy with from time to time is that economics' aspirations to the status of "hard" science are fostered in no small measure by thre existence of the Nobel Prize for Economics. I've lost count of the number of times in internet discussions or media pieces I've encoutered references to some economist having won the Nobel, as if that were somehow evidence of the correctness of his theories. The problem with that is that, as far as I can make out, the Nobel Prize for Economics is awarded on the basis of a hypothesis or model being both novel and coherently formulated and argued, but not on the basis of how well that model applies to reality. Consider the fact that both John Maynard Keynes and Milton Friedman received the Nobel Prize for Economics, in spite of the fact that their theories are clearly mutually contradictory.

— • —

[In the deluge of spam that we just received, we accidentally deleted Jurjen's comment, and had to re-enter it manually. Accordingly, the date- and time-stamp are now incorrect. The original comment was posted on 27th June at 6:19AM EDT. Please pardon our error — the Management.]

I agree that many economists do lose sight of the assumptions that they are making and the impact of various other circumstances. Indeed, if I recall correctly, 'twas my [user]namesake who helped start model economics after physics, hugely changing the field and its progression til now. However, economists have been increasingly taking into account all of these other behavioural factors for quite a long time. Multiple chains of this have been happening for the better part of a century, half a century, and 30-40 years. And it is certainly time for interdisciplinarity to reign as it has been doing recently, something I am involved in and hope to be more so.

Also, Keynes never won a Nobel in Economics. Don't know where you got that from. But even if Keynes had gotten one, it wouldn't bode ill; advancements can occur even through false ideas (more specifically, ideas which are later found to be false).

Something I've always wondered about: why does a field not being a "hard science" (or "natural science" as I call them) make it less scientific? Perhaps the methods of some social "scientists" are not very scientific, but that's different from the field not being a natural science.

The results of the hard sciences are more replicable, so the science is more certain. Soft sciences are kinda squishy--the results are far more tentative because the results are so hard to replicate. Given that, the soft sciences seem more prone to wandering off into the weeds for extended periods.

I'm not saying that the hard sciences don't amble down dead ends (n-rays, anyone?), but really, since Copernicus turned the heavens inside out, all the so-called revolutions in physics have been more along the lines of refinements. You don't need Einstein to get to the moon, but you probably do to get to Alpha Centauri. Same since Darwin in biology. The discovery of the structure of DNA unlocked a lot of doors, but really, it just finally gave a physical mechanism to "descent with modification".

I don't think you can say the same with respect to economics or anthropology. Mainly because it's impossible to run the same experiment twice in those fields.

Replicability certainly occurs in the social sciences. Experiments are conducted all the time in psychology or economics. True, often you can't conduct larger experiments because of complexity (due to the presence of humans) or ethics (also due to the presence of humans...), although there are natural experiments. I certainly agree that conclusions are much more tentative because humans are very complex. But humans exist and display objective phenomena, even though these phenomena can be extraordinarily complex.

Maybe it's just semantics, but I dislike the connotations (or perhaps denotation) involved in calling a social science less scientific. The way individuals in that field conduct themselves may definitely be un- or pseudo-scientific, but calling the field unscientific feels to me like we're writing it off, despite good insights which come from it. It really hinders progress when trying to take these real and scientific insights and apply them to other areas.

I'm not trying to say that the soft sciences are worthless, though maybe it came across that way. It's just that there are so, so many more variables in the soft sciences, and you can rarely hold many of them constant for very long, if at all. It really is a lot harder to replicate results in the soft sciences. That doesn't mean we should give up on the scientific approach in any field, just that we have to regard the results as more tentative, less reliable, less predictive in some areas.

As far as I'm concerned, science is the only way we know things. Since people are part of the observable universe, the only way we're going to understand them is with a scientific approach. Unfortunately, people being people, the occasional wild hair crawls up the occasional ass, and then all bets are off. (My theory--it aggravates the ass gnomes.)